Approaches to Improve Student Feedback with Wisdom of Crowd
Ms. P. V Wazalwar, Ms. M. A Potey · International Journal of Computer Trends and Technology · 2014
The rapid development in Crowdsourcing offers new ideas to the traditional methods for improving Information Retrieval systems in design, training and evaluation. Recently, Crowdsourcing has been used as a source of relevant and non- relevant judgements. Crowdsourcing is an outsourcing a task to crowd/people where all the steps from design to implementation are included. The popularity of Crowdsourcing is increasing rapidly because of the large number of crowd involvement. The purpose is to apply wisdom of crowd strategy in academia. For this purpose we provide student feedback system, in which feedbacks can be both relevant and non-relevant. After that feedbacks are pre-process. The feedbacks are then clustered for the classification of feedbacks using similarity based technique. Our system takes related feedbacks. We considered two approaches k-means and hierarchical algorithm to implement a text mining methodology. We have used hierarchical clustering to classify student feedbacks and compare our work with the k- means approach for clustering and evaluated our approach with standard metrics precision, recall and F1-Score.